A Hybrid Optimization Driven Deep Residual Network for Sybil Attack Detection and Avoidance in Wireless Sensor Networks

Small Sensor nodes that can sense, store, and send data make up a wireless sensor network (WSN). In addition, the sensor nodes are limited when it comes to energy along with computing capability. Clustering is a good way to reduce the amount of energy used by sensor nodes when transferring information to the ground station (GS) from the transmitter node. For information transmitting among nodes, multi-path routing is utilized. The WSN, on the other hand, is vulnerable to a variety of security vulnerabilities that might decrease network performance. WSNs are vulnerable to a variety of assaults, which have harmed the network’s overall performance. Consequently, the fundamental objective of this study is to design and create a WSN-based hazard detection and avoidance systems. This research proposes a learning technique towards warning systems and defense in Wireless Sensor Networks that is both energy-efficient and optimization-aware. This method begins with the emulation of WSN nodes. The suggested model’s overall methodology includes WSN simulation, selection of cluster heads forwarding to BS, Finally, BS has a Sybil approach to detect and assault protection. Following the WSN simulation, the LEACH Technique is used to select its most sustainable power cluster head. Such information would then be transported using the fractionated artificial bee colony (FABC) method. After the data is collected at BS, it is used to execute assault mitigation and detection activities. The Jaro-Winkler gap is employed to identify the essential feature in data. Following the identification of the ideal rules, The Sybil assault use the deep residual network (DRN). To train DRN, the Competitive Multi-Verse Rider Optimizer (CMVRO) was suggested, which blends the Competitive Multi-Verse Optimizer (CMVO) with both the Rider Optimization Algorithm (ROA). This bit rate is employed to counteract Sybil’s assault. In this case, the throughput is used.